Skip to content
beautydew.
Hair & scalp

Coconut Oil and Hair: Who's Actually Done the Research?

"Oil your hair before you shampoo" usually points to one finding: coconut oil, unlike mineral or sunflower oil, cut protein loss during washing in lab testing. What rarely gets mentioned is that every paper behind that claim, across twenty years, came out of the R&D lab of the company that makes India's top-selling coconut hair oil — and no one outside it has published a repeat.

Does coconut oil actually protect hair during a wash, or is that just marketing?

Short answer: The lab measurements are real. In controlled testing, coconut oil reduced the protein hair loses during washing, while mineral oil and sunflower oil didn't. But nearly every study behind that finding, across three papers and twenty years, was run by the R&D lab of the one company that sells India's best-known coconut hair oil — and no outside lab appears to have repeated it.

What the lab actually found

The most-cited comparison tested coconut, mineral, and sunflower oil on hair tresses, some undamaged, some pre-damaged, oiled before or after washing, then measured protein loss afterward, since that's one way washing wears hair down. Coconut oil was the only oil that reduced it, damaged or not, either way applied; mineral and sunflower oil showed no effect. [1] A specific percentage circulates online but can't be confirmed from an accessible text, so it isn't repeated here.

A 2005 follow-up used a different method: force needed to pull an oiled fiber free, and how a thinning oil film changes reflected light, a proxy for oil soaking in rather than sitting on top. Coconut oil's film thinned over 24 hours, then roughly 40% more with a few minutes of heat; mineral oil barely moved. [2] The likely reason is molecular shape: lauric acid, coconut oil's main fat, is small and binds well to keratin; mineral oil has no such pull, and sunflower oil's kinked molecule is bulkier and harder to thread through. [1][2] Both used tresses and fibers, not people: a mechanism, not a clinical trial.

One company, three papers, no independent check

Trace who ran these studies and it's the same company each time. The 2003 study's authors worked in Marico Industries Ltd.'s R&D department. The 2005 follow-up's materials were supplied by Marico too, same affiliation on the byline. A 2021 repeat, with newer methods, also came from Marico's R&D center. [1][2] Two decades, three papers, one company — no outside lab has repeated the test.

That doesn't make it wrong. Companies test their own materials constantly, and plenty of that work holds up — drug approvals lean on industry-funded trials more than people assume, checked by regulators rather than dismissed. The methods here look like standard cosmetic science, and the numbers stay consistent across both papers. [1][2] What's missing is the check that lets you fully trust it: nobody without a stake has independently repeated this. That's not an accusation of cherry-picked data — there's no evidence of that — only that a single source is weaker than a confirmed one. Measured, plausible, unconfirmed is the honest read; not proven, not fake.

Should you still oil your hair before washing?

Probably fine to try, with realistic expectations. Coconut oil is cheap, low-risk for most people, and the mechanism is reasonable even without outside confirmation. It won't undo damage already done — hair past the scalp isn't living tissue, so no oil rebuilds a broken cuticle. Conditioners work the same temporary way, smoothing the strand until the next wash. [3] Nothing points to faster growth, which happens at the follicle, not the strand.

Separately: coconut oil is rich in lauric acid, a fatty acid Malassezia, the yeast behind fungal acne and scalp bumps, can use as food in lab studies, though a comparatively weak trigger next to palmitic and oleic acid. [4][5] Already own an oil-heavy product? Run the label through our ingredient scanner; prone to breakouts along the hairline or scalp? Our fungal acne checker is a faster first stop.

FAQ

Does industry funding mean the coconut oil result is fake?

No. It means nobody outside the company selling a product built on this claim has checked it. The measurements may be accurate — tress testing is a standard, checkable method — but evidence measured once, by an interested party, never independently repeated, is weaker than a confirmed result. Neither is the same as fabricated.

Does this mean other oils don't do anything for hair?

No. This specific effect wasn't shown for mineral or sunflower oil here. Oils generally can still smooth the cuticle and ease combing by coating the surface; coconut oil's proposed advantage goes beyond that, into the fiber.

Will pre-wash oiling make my hair grow faster or thicker?

No. Growth happens at the follicle, under the scalp, which a topical oil never reaches. This is about protein loss during washing, a maintenance effect, not a growth effect.

Is coconut oil safe to use on the scalp?

For most people, yes, though not universally. It contains a fatty acid Malassezia can use as fuel in lab conditions, but a weak one next to others. [4][5] If your scalp is prone to fungal acne or breakouts, check your products first, and see a dermatologist for anything sudden, painful, or spreading.

References

  1. Effect of mineral oil, sunflower oil, and coconut oil on prevention of hair damageJournal of Cosmetic Science, 2003
  2. Investigation of penetration abilities of various oils into human hair fibersJournal of Cosmetic Science, 2005
  3. Shampoo and Conditioners: What a Dermatologist Should Know?Indian Journal of Dermatology, 2015
  4. Fatty-acid metabolism of Pityrosporum ovaleBiochemical Journal, 1968
  5. Lipid-dependent growth of Malassezia species in defined medium with single free fatty acids (C12–C24)FEMS Yeast Research, 2025

The occasional email: new launches, price drops, one decoded ingredient breakdown, and upcoming beauty events. No spam, no data sold — unsubscribe in one click.

Command palette

Search products and ingredients, or jump to anywhere in beautydew labs.